Portable AI memory is your context as a thing you own and carry, rather than a profile each AI company keeps about you inside its walls. You hold one file of who you are, any assistant reads it with your consent, and it leaves when you say so.
Every major lab launched memory in the last year, and every one of them built it as a silo. That is not laziness. The more an assistant knows you, the harder it is to leave, so your accumulated context is the retention strategy. Asking a lab for portable memory is asking it to make quitting easier. The incentive runs the other way, permanently.
Three things. Ownership: you can read and edit every byte, not a summary. Portability: it works by paste or a standard grant in any AI, not through one vendor’s API. Revocation: you can end all access at once and prove it ended, even offline. Miss any one of the three and it collapses back into somebody’s product feature.
You take a short free read, and it becomes a signed file: your word, your rhythm, how to meet you. You add it to ChatGPT or Claude in about two minutes, curate exactly which lines travel, and one tap revokes everything, verifiable against a public status list. It is small, honest, and independent, which is roughly what a thing like this has to be.
Take the free four-minute read, get your word and your one-of-a-kind eye, and add your pupil to ChatGPT or Claude in about two minutes. Revoke it all with one tap, anytime.
Take your read, free Watch the revoke demoNo. Exports give you a zip of the past. Portable memory is a living file the next AI can actually use on message one, and that you can take back.
The opposite, done right. Instead of a dozen companies each keeping an opaque profile, one consented file travels, you see every read, and revocation is provable.